What representation do you prefer for images, 2 dimensional or 1 dimensional?

In General, 2D is more common in image processing and the majorty of methods uses 2D (i.e. Filters, wavelet, curvelet, detection and so on)

2 dimensional because image can't be 1D

2D is relatively better option, and 3D would provide more depth information as well

Absolutely 2 dimensional :)

Image can't be in 1-D... as image is stored in the form of pixel value and each image has some width and height. So if the pixels are in Grayscale (Black & White) form then we can use 2-D  (in which each value represents a pixels gray level) but if the image is in colored form the it should be represented in 3-D form one for Red color one for Green and one for Blue

Best way to represent the images is three dimensional.

in my openion i think that the image who have a 2 dimensionale is best than the other

i thing its became N*2 to N*4 dimensional.

Generally images are represented by 2D array. But for fast processing, some matlab functions like colfilt convert images to 1D by concatinating image columns. This Function process images faster than other that uses 2D representation for images.

2 dimensional representation is more appropriate for image processing